Object recognition or identifying a region of interest in an image are of great importance in numerous fields and applications in various industries. For example, detecting or identifying a person in an image may be used to extract the image of the person, replace the background and/or apply effects such as matting, background smoothing etc.
Image acquisition devices such as digital cameras are now widely available, for example, most mobile phones are now equipped with a camera. However, current systems and methods for object recognition in an image rely on extensive processing of the image. For example, in order to recognize specific objects in an image, a digital image may need to be uploaded to a server or provided to a home computer where an image processing application performs identification of regions of interest or object recognition.